egies of mixed development as an alternative way of using the forest for human needs. They claim that intensive, stable agricultural land use with a mixture of crops and livestock can be combined with labor-intensive efforts to maintain soil quality by careful, thorough tillage, agroforestry, manuring, terracing, irrigation, and drainage. Thus they can provide high, reliable, sustainable production from smallholdings with high inputs of household labor and little capital or fossil fuel energy. These systems may also help preserve mature forest ecosystems from destruction by reducing development pressure on them (Anderson, 1990).
The potential for a future of less-extensive forest use in the Amazon Basin relates in part to land distribution. Inequality of land holdings in Brazil has increased greatly over the last few decades, with 70 percent of Brazilian farmers now landless and 81 percent of the farmland held by just 4.5 percent of the population (Hildyard, 1990). This pattern of increasing inequality also holds in the Amazon, making access to resources more difficult for subsistence farmers and hunters and gatherers and threatening indigenous land tenure systems based on communal rights (Chernela, n.d.; Poole, 1989). Larger landholdings bring more extensive use. In Pará state, for instance, small farms cultivate an average of 50 percent of their area, while farms of over 1,000 ha cultivate only 26 percent (Hecht, 1981). More intensive cultivation means that less forest must be displaced to meet human needs. Moreover, stable smallholders have an incentive to economize on land and keep it productive, so that land degradation can be slower with more intensive use. Thus, the current pattern of extensive development, by displacing indigenous peoples and small-scale extractors, has removed a brake on deforestation and threatens a store of valuable knowledge about the intensive management of forest species for human consumption.
There are barriers to a transition to a mixed-development strategy in the Amazon. One is the social change resulting from the current extensive strategy. Another is the politics of change. With rural poverty increasing and a political choice between dividing up large landholdings and encouraging the landless to colonize unclaimed or "unused" frontier lands, migration and resettlement policies are much the more palatable alternative (Macdonald, 1981). And finally, there are intrinsic social limits. Although portions of the local environment could support intensive land use like that of the wet rice/garden systems of south China and Java, the necessary local density of population with plentiful labor and nearby markets are not present (Moran 1987:75). Extensive, extractive